On September 8th, the Fifth Circuit Court of Appeals of the United States partially upheld a decision that found several public officials had coerced social media companies into censoring speech protected by the First Amendment. This was done through veiled threats and subtle forms of intimidation made by administrative officials in emails, calls, and regular meetings with corporate executives. The plaintiffs—composed of private parties who felt censored by platforms and two states—argued that behind the supposedly private moderation decisions laid the hand of the federal government.
Americans call this area of the law jawboning, in reference to the jawbone that is moved when we talk, which is the mechanism through which pressures of these sorts are exerted. It is an extremely complex area of law, in part because distinguishing when public officials cross that fuzzy legal line depends on assessing the nature actions that happen in private settings in light of vague and ambiguous criteria. In this piece, I explain why the occurrence of jawboning might be an inevitable feature of modern administrative governance, and outline both the unique challenge that underpins any attempt to legally regulate it as well as the urgency of doing so.
Jawboning: An Inevitability of Administrative Governance?
Biden v. Missouri turned on the allegedly coercive nature of a set of communications between public officials and executives at tech companies. These communications, often in the form of emails, showed a relationship of collaboration and tension, in which public officials made requests and corporate officials provided explanations, excuses, and information. The record reveals a persistent and frustrated group of public officials who make content moderation demands and ask harshly for policy changes. Corporations appear compliant and submissive.
The Fifth Circuit held that while urging, pressuring, or even inducing a private actor to censor protected speech does not violate the Constitution, coercing or significantly encouraging them to do so does. In holding that the latter had occurred, the Court focused on the regulator’s “direct involvement” in the decision-making process (p. 34) as well as their word choice and tone, whether the speech was “perceived” as a threat or not, whether there is regulatory authority vested on the speaker, and whether adverse consequences were foreseen (p. 44).
At times, jawboning cases seem like hair-splitting exercises. For instance, in Biden v Missouri the Fifth Circuit found that the line that separates permissible speech from illicit threats had been crossed, but in O’Handley v. Weber and Twitter– in a case about a collaborative “trusted flagger” relationship between Twitter and California state officials-the Court found the collaboration was voluntary and that the company executives in O’Handley acted by their own free will.
The decisions evidence the difficulty of establishing workable legal tests to regulate informal processes of policy-making that are arguably both an unavoidable and even necessary part of our system of governance. They are unavoidable for the rule-producing stage of politics is only the final step of a process that starts well before a statute is adopted. A statute, before being a statute, is a bill, and a bill—before being a bill—is often a hearing, a request made to a legislator by concerned constituents, a problem named in the public sphere, and so on. In the administrative state, where enforcement and rule-making adopt different forms, the policy-making process is even more complex and—on occasions—invisible. Informal calls, emails, meetings, speeches delivered in not-too-public settings, are all ways through which public officials communicate with individuals and corporations subjected to their jurisdiction.
It is within these informal settings and communications that the risk of jawboning arises. Courts have tried to distinguish between what are legitimate assertions of regulatory authority and jawboning by focusing- among other factors- on a regulator’s capacity to actually follow up their denied requests with formal rule-change. Yet, this ignores that a regulator’s informal remarks will likely always be perceived as coming from someone with such capacity, irrespective of whether the particular regulatory actor in question actually holds formal rule-making power. Due to this “cloak of authority” that surrounds regulators and administrative officials, the risk of jawboning occurring is basically inevitable,
Notably, contrary to the formal processes of rule making, the informal channels of communication through which jawboning occurs cannot be subjected to the values of transparency, accountability, participation, and stakeholder engagement that we ordinarily insist upon. This is not to be taken lightly. We insist on these values because they capture core democratic values implicit in the design of our representative institutions. We want to be involved in decision-making processes and we want to see how decisions are made, in part to ensure that our votes, voices and interests are considered and respected by those who represent us. Yet, despite their incompatibility with these requirements, the informal processes of policy making that give rise to the possibility of jawboning appear to be not just an unavoidable but also a necessary aspect of modern administrative governance. In particular, spaces for sharing insights between public and corporate officers seem essential to make informed public policy. These spaces cannot be totally formalized, nor can they be made absolutely transparent. But while they must exist, they also pose the problems we are discussing.
The Limits of Line Drawing
The issue with regulating jawboning is thus rooted in a paradox: we want our cake and eat it too. Indeed, we want our political process to be open and transparent and subjected to contestation by different stakeholders, but we also want our public officials and administrators to be effective in what they do. This means allowing them to communicate with individuals and corporations in informal settings in which they can make requests with the purpose of guiding private conduct in a way that complies with public policy goals established by law. And this includes discussing and imagining possible regulatory solutions to deal with specific problems.
This paradox explains why the law deals with jawboning through line drawing questions: it is ultimately a question of when public officials went too far. But if the problems raised in the previous section are sound, the line drawing exercise hardly solves them. Courts have proposed standards that are difficult to administer from the point of view of public officials. In Vello, the New York public official who signaled that she was willing to show leniency in her enforcement obligations if certain conduct change happened was found to be moving within the boundaries of permissible government speech. But in Missouri v. Biden the official mentioning that in the White House they were “considering our options” was found out of line. Perhaps it is the consistency of a pattern of interactions that becomes unacceptable, but this does not come out clearly from the case law. This leaves us not with a hard-coded rule but a vague, general standard calling for caution that carries the risk of unduly obstructing effective regulatory action.
To make things worse, to determine whether a particular set of interactions went too far requires a level of evidence that can only be produced through judicial discovery. Yet, cases remain rare, not least because corporate officials generally have little incentive to undermine their channels of communication with public officials: they usually take advantage of them for lobbying purposes. As a result, jawboning cases are few when compared to the extent to which informal communications are used to make policy. Only those who feel injured by the effects of these informal policy-making mechanisms end up pursuing legal action, whether it was the Writer’s Guild affected by the Family Hour adopted by the National Association of Broadcasters, the authors of the books being questioned by Senator Warren, or the physicians, activists, and Attorney Generals who felt censored for questioning Covid-19 federal policies in Missouri v. Biden.
Jawboning as A Global Regulatory Phenomenon?
Pressuring corporations is a normal governance mechanism that has been found e.g. in use in Spain to deal with utility companies, in Sweden to deal with the local textile industry, and in India and Canada to deal with the governance of the Internet.
A newly launched research project at CELE has similarly found that electoral authorities in Mexico, Brazil, and Argentina asked corporations to sit down at informal meetings with the purpose of creating fluid channels of communication between public and corporate officials, especially around election season and to deal with disinformation. In the three countries all major Internet platforms voluntarily joined.
Outside the US, however, there is almost no case law on how to legally regulate the appropriate scope of these practices. This might be due, not least, because of the need for efficient judicial discovery to establish the factual patterns required to establish jawboning. Judicial inquiries in other jurisdictions rarely produce similar levels of detailed paper trails as occurs in the US, rendering the occurrence of jawboning not only be difficult to prove but practically invisible. To make matters worse, states also enjoy different levels of regulatory power. The power a US public official to pressure Internet corporations into complying with her desires is different from the power enjoyed by peers in or with weaker market-shared from the point of view of companies’ business models.
How to Regulate the Regulators
A comparative research agenda on jawboning is thus challenging. Yet, its pursuit is arguably of particular urgency, not least given that the exertion of informal regulatory pressures is one of the main mechanisms through which the Internet is governed. The informal streamlined channels of communications between corporate and public officials are currently made possible by existing protections against intermediary liability. However, this is changing. Thus, the EU’s Digital Services Act (DSA) is an attempt to formally regulate these relationships: there are specific obligations, oversight mechanisms, and potential sanctions in the future of the DSA’s implementation. Transparency laws at the state level in the United States also move towards more routinized interactions.
In this context of incipient global regulatory change an opportunity arises to think more deeply about how to effectively regulate the regulators. The US case law on jawboning provides a useful starting point in this regard but research from other jurisdictions is necessary to develop adequate criteria. One thing the Missouri v. Biden saga showed is the importance of creating adequate paper trails of all communications between public and corporate officials. As such, forthcoming regulations should develop paper-trail obligations in the context of more routinized interactions. It seems like a necessary condition to draw the difficult line that separates effective government from regulatory bullying.